1. Field of the Invention
The present invention relates to an information processing apparatus, a control method for the same, a program, and a storage medium. More particularly, the present invention relates to an information processing apparatus and a control method for the same, which can set an information extraction condition and an information exclusion condition with a UI operation input by a user for icon data. Further, the present invention relates to a program and a storage medium for use with the information processing apparatus and the control method.
2. Description of the Related Art
Hitherto, in an information extraction apparatus configured to extract desired data from among a lot of data, necessary information has been extracted by designating an information keyword which is directly included in the desired data, or by designating an information keyword which is included in various metadata associated with the desired data.
When the information keyword is directly included in the desired data, it is given as an arbitrary keyword. Also, when the information keyword is included in various metadata, it is given as an arbitrary or preset keyword depending on the kind of metadata designated.
Further, when a plurality of information keywords are designated from among many information keywords in the metadata, information extraction with point rating is also practiced in consideration of, e.g., weights assigned to the individual information keywords. Thus, in order for a user to extract information from a lot of data with high efficiency, it is important to perform proper selection of the information keyword, proper classification of information, and proper weight assignment.
The following several techniques have been disclosed so far to allow the user to perform the proper selection of the information keyword, the proper classification of information from an optimum viewpoint, and the proper weight assignment.
Japanese Patent Laid-Open No. 2003-345829 proposes a technique of automatically extracting labels which are used to narrow down an initial search result, and presenting the labels to users.
Even when a user is not previously acquainted with information regarding what labels serve as information keywords, the user can efficiently narrow down the search by selecting the proper one of the labels. Also, by repeating proper selection of the label, the user can obtain the desired information.
Japanese Patent Laid-Open No. 08-044759 proposes a technique as follows. When a user views data that is a target of retrieval, the data is displayed in accordance with a classification viewpoint which is previously prepared depending on the data. At that time, when the user changes the classification viewpoint from one to another, the classified search result is displayed in a dynamically changing way depending on respective viewpoints.
Accordingly, the user can narrow down the target data in a short time by efficiently selecting the proper viewpoint while successively changing the viewpoints. In addition, the user can be provided with a new viewpoint that has not been taken or managed by the user up to that time.
Japanese Patent Laid-Open No. 08-339375 proposes a technique as follows. When a retrieval request item is input as a retrieval keyword, retrieval target items are selected based on comparison between the input retrieval request item and the retrieval target items and are displayed such that the retrieval target items are arrayed with the closest one, in terms of the degree of its similarity or similarities to the retrieval request item, located at a center.
Further, when one retrieval target item which is regarded as an unnecessary retrieval result is moved away from the closest retrieval target item located at the center, other items having similarities relatively closer to the retrieval target item that has been moved away are also moved away from the center in ways that depend on their respective similarities to the moved retrieval target item. Consequently, unnecessary retrieval results are displayed farther away from the closest retrieval target item, thus allowing the user to more easily select one or more objective items only.
Meanwhile, at the homes of today, the amount of image content data is rapidly increasing and the demand for extraction of desired information from the great amount of image content data available is also increasing at the same time. The image content data is sufficiently associated with metadata for information extraction, such as shooting data/time and shooting equipment.
When a user finally extracts the objective image content data, the extraction is often performed by utilizing not only the above-described information keywords, but also thumbnail information representing a summary of the content, i.e., a visual search by the user.
In the above-described information extraction apparatus, however, proper extraction of the information demanded by the user cannot be realized unless proper information keywords are selected and appropriately combined with each other. Further, even when information is extracted, a very large number of hits may result in many cases. In addition, if information keywords are erroneously selected, the objective information demanded by the user cannot be obtained in many cases even when the information is extracted in proper number.
With the technique described in Japanese Patent Laid-Open No. 2003-345829, labels presentable to the user are automatically extracted. However, selection of the label fails, for example, when the relationship between the respective labels and data corresponding thereto is not definite to the user, or when a large number of labels are presented to the user. Another problem is that the automatically extracted label corresponds simply to the presence or absence of the information keyword, and a parameter corresponding to a metadata item cannot be designated.
With the technique described in Japanese Patent Laid-Open No. 08-044759, when the user does not have a definite viewpoint for narrowing down the target data, data extraction based on narrowing-down cannot be efficiently performed and the objective data cannot be extracted in some cases. As with the foregoing technique, another problem is that the presented viewpoint corresponds to the presence or absence of the information keyword and a parameter corresponding to a metadata item cannot be designated.
With the technique described in Japanese Patent Laid-Open No. 08-339375, the result of a user's operation of moving one item is given as movements of other associated items depending on the similarities, which however are unknown to the user. Also, the distances through which the other items are moved are determined such that the user can more easily view those items. In other words, the user's operation of moving the item, although it affects the location of other items, is still not a meaningful operation such as designating a parameter, in terms of the user being able to exercise a high degree of control of the search by this means.
Another problem is that, because the item positions currently displayed depend on the similarities, the items are moved in directions in a way that may differ from the feeling of the user who has moved the item. Further, when the user visually searches, e.g., image contents while viewing thumbnails, the visual search is very inefficient if there are a large number of thumbnails.